tensorflow67 《深度学习原理与TensorFlow实战》04 CNN看懂世界 04深度残差网络

00 环境

#《深度学习原理与TensorFlow实战》04 CNN看懂世界
# 书源码地址:https://github.com/DeepVisionTeam/TensorFlowBook.git
# 视频讲座地址：http://edu.csdn.net/course/detail/5222
# win10 Tensorflow1.2.0 python3.6.1
# CUDA v8.0 cudnn-8.0-windows10-x64-v5.1
# https://github.com/tflearn/tflearn/blob/master/examples/images/residual_network_mnist.py # https://github.com/tflearn/tflearn/blob/master/examples/images/residual_network_cifar10.py[/code] 
01 residual_network_mnist.py
# -*- coding: utf-8 -*-

""" Deep Residual Network.

Applying a Deep Residual Network to MNIST Dataset classification task.

References:
- K. He, X. Zhang, S. Ren, and J. Sun. Deep Residual Learning for Image
Recognition, 2015.
- Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner. "Gradient-based
learning applied to document recognition." Proceedings of the IEEE,
86(11):2278-2324, November 1998.

Links:
- [Deep Residual Network](http://arxiv.org/pdf/1512.03385.pdf)
- [MNIST Dataset](http://yann.lecun.com/exdb/mnist/)

"""

from __future__ import division, print_function, absolute_import

import tflearn
import tflearn.data_utils as du

# Data loading and preprocessing
import tflearn.datasets.mnist as mnist
X, Y, testX, testY = mnist.load_data(one_hot=True)
X = X.reshape([-1, 28, 28, 1])
testX = testX.reshape([-1, 28, 28, 1])
X, mean = du.featurewise_zero_center(X)
testX = du.featurewise_zero_center(testX, mean)

# Building Residual Network
net = tflearn.input_data(shape=[None, 28, 28, 1])
net = tflearn.conv_2d(net, 64, 3, activation='relu', bias=False)
# Residual blocks
net = tflearn.residual_bottleneck(net, 3, 16, 64)
net = tflearn.residual_bottleneck(net, 1, 32, 128, downsample=True)
net = tflearn.residual_bottleneck(net, 2, 32, 128)
net = tflearn.residual_bottleneck(net, 1, 64, 256, downsample=True)
net = tflearn.residual_bottleneck(net, 2, 64, 256)
net = tflearn.batch_normalization(net)
net = tflearn.activation(net, 'relu')
net = tflearn.global_avg_pool(net)
# Regression
net = tflearn.fully_connected(net, 10, activation='softmax')
net = tflearn.regression(net, optimizer='momentum',
loss='categorical_crossentropy',
learning_rate=0.1)
# Training
model = tflearn.DNN(net, checkpoint_path='model_resnet_mnist',
max_checkpoints=10, tensorboard_verbose=0)
model.fit(X, Y, n_epoch=100, validation_set=(testX, testY),
show_metric=True, batch_size=256, run_id='resnet_mnist')
'''
Extracting mnist/train-images-idx3-ubyte.gz
Extracting mnist/train-labels-idx1-ubyte.gz
Extracting mnist/t10k-images-idx3-ubyte.gz
Extracting mnist/t10k-labels-idx1-ubyte.gz
---------------------------------
Run id: resnet_mnist
Log directory: /tmp/tflearn_logs/
---------------------------------
Training samples: 55000
Validation samples: 10000
--
Training Step: 1  | time: 6.395s
| Momentum | epoch: 001 | loss: 0.00000 - acc: 0.0000 -- iter: 00256/55000
Training Step: 2  | total loss: 2.07091 | time: 11.896s
| Momentum | epoch: 001 | loss: 2.07091 - acc: 0.0844 -- iter: 00512/55000
Training Step: 3  | total loss: 2.25693 | time: 17.365s
...
Training Step: 242  | total loss: 0.07294 | time: 156.351s
| Momentum | epoch: 002 | loss: 0.07294 - acc: 0.9785 -- iter: 06912/55000
Training Step: 243  | total loss: 0.07384 | time: 162.524s
| Momentum | epoch: 002 | loss: 0.07384 - acc: 0.9783 -- iter: 07168/55000
Training Step: 244  | total loss: 0.07520 | time: 168.575s
| Momentum | epoch: 002 | loss: 0.07520 - acc: 0.9789 -- iter: 07424/55000
...
'''


02 residual_network_cifar10.py
# -*- coding: utf-8 -*-

""" Deep Residual Network.

Applying a Deep Residual Network to CIFAR-10 Dataset classification task.

References:
- K. He, X. Zhang, S. Ren, and J. Sun. Deep Residual Learning for Image
Recognition, 2015.
- Learning Multiple Layers of Features from Tiny Images, A. Krizhevsky, 2009.

Links:
- [Deep Residual Network](http://arxiv.org/pdf/1512.03385.pdf)
- [CIFAR-10 Dataset](https://www.cs.toronto.edu/~kriz/cifar.html)

"""

from __future__ import division, print_function, absolute_import

import tflearn

# Residual blocks
# 32 layers: n=5, 56 layers: n=9, 110 layers: n=18
n = 5

# Data loading
from tflearn.datasets import cifar10
(X, Y), (testX, testY) = cifar10.load_data()
Y = tflearn.data_utils.to_categorical(Y, 10)
testY = tflearn.data_utils.to_categorical(testY, 10)

# Real-time data preprocessing
img_prep = tflearn.ImagePreprocessing()
img_prep.add_featurewise_zero_center(per_channel=True)

# Real-time data augmentation
img_aug = tflearn.ImageAugmentation()
img_aug.add_random_flip_leftright()
img_aug.add_random_crop([32, 32], padding=4)

# Building Residual Network
net = tflearn.input_data(shape=[None, 32, 32, 3],
data_preprocessing=img_prep,
data_augmentation=img_aug)
net = tflearn.conv_2d(net, 16, 3, regularizer='L2', weight_decay=0.0001)
net = tflearn.residual_block(net, n, 16)
net = tflearn.residual_block(net, 1, 32, downsample=True)
net = tflearn.residual_block(net, n-1, 32)
net = tflearn.residual_block(net, 1, 64, downsample=True)
net = tflearn.residual_block(net, n-1, 64)
net = tflearn.batch_normalization(net)
net = tflearn.activation(net, 'relu')
net = tflearn.global_avg_pool(net)
# Regression
net = tflearn.fully_connected(net, 10, activation='softmax')
mom = tflearn.Momentum(0.1, lr_decay=0.1, decay_step=32000, staircase=True)
net = tflearn.regression(net, optimizer=mom,
loss='categorical_crossentropy')
# Training
model = tflearn.DNN(net, checkpoint_path='model_resnet_cifar10',
max_checkpoints=10, tensorboard_verbose=0,
clip_gradients=0.)

model.fit(X, Y, n_epoch=200, validation_set=(testX, testY),
snapshot_epoch=False, snapshot_step=500,
show_metric=True, batch_size=128, shuffle=True,
run_id='resnet_cifar10')
'''
# 注意自解压目录，需要手动修改一下目录层次，cifar-10-batches-py==>cifar-10-batches-py/cifar-10-batches-py。
---------------------------------
Run id: resnet_cifar10
Log directory: /tmp/tflearn_logs/
---------------------------------
Preprocessing... Calculating mean over all dataset (this may take long)...
Mean: [ 0.49139968  0.48215841  0.44653091] (To avoid repetitive computation, add it to argument 'mean' of `add_featurewise_zero_center`)
---------------------------------
Training samples: 50000
Validation samples: 10000
--
Training Step: 1  | time: 3.993s
| Momentum | epoch: 001 | loss: 0.00000 - acc: 0.0000 -- iter: 00128/50000
Training Step: 2  | total loss: 2.08295 | time: 7.453s
| Momentum | epoch: 001 | loss: 2.08295 - acc: 0.0562 -- iter: 00256/50000
Training Step: 3  | total loss: 2.26373 | time: 10.725s
...
Training Step: 121  | total loss: 1.72538 | time: 425.664s
| Momentum | epoch: 001 | loss: 1.72538 - acc: 0.3421 -- iter: 15488/50000
Training Step: 122  | total loss: 1.72287 | time: 429.461s
| Momentum | epoch: 001 | loss: 1.72287 - acc: 0.3415 -- iter: 15616/50000
Training Step: 123  | total loss: 1.71641 | time: 433.079s
| Momentum | epoch: 001 | loss: 1.71641 - acc: 0.3448 -- iter: 15744/50000
...
'''